browser-qa

Pass

Audited by Gen Agent Trust Hub on Aug 25, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes data from external sources and local files that may contain instructions targeting the agent, creating a potential path for indirect prompt injection.
  • Ingestion points: In SKILL.md (Step 2), the agent is instructed to fetch content from issue trackers (GitHub/GitLab) and files located in docs/, specs/, or .scratch/ directories.
  • Boundary markers: The instructions do not specify any delimiters or safety prompts to treat the fetched content as untrusted data rather than instructions.
  • Capability inventory: In SKILL.md (Step 3), the agent uses the agent-browser skill to drive web flows, providing a mechanism where injected instructions could potentially interact with web applications or capture data.
  • Sanitization: There are no mentioned mechanisms for escaping, validating, or filtering the content retrieved from issue trackers or specification files.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 25, 2026, 12:34 PM
Security Audit — agent-trust-hub — browser-qa